Primary prevention of HIV is as important today as it was in the early years of the AIDS epidemic. Evaluating behavioral interventions has been difficult because self-reported measures of sensitive and illegal behaviors associated with HIV are vulnerable to nontrivial levels of reporting bias. The proposed research seeks to improve the investigators' understanding of bias in self-reported measures in a population at very high risk for HIV, namely STD clinic patients. Data from recent population-based surveys have found that respondents assigned to a new audio computer-assisted self-interview (ACASI) system are more likely to report engaging in a range of sexual behaviors and illicit drug use than respondents assigned to interviewer administered questionnaires (IAQs) or paper and pencil self-administered questionnaires. However, a pilot study of ACASI in a small sample of STD clinic patients, found limited mode effect for even the most sensitive items. Further examination of the data indicate that reporting patterns of new patients are different from that of repeat patients. From these data, the investigators hypothesize that (1) interview mode effect for sensitive measures of sex and drug use will be more pronounced among new patients than repeaters, but (2) mode effect for measures of prescribed behaviors (i.e., condom use) will be more pronounced among repeaters. The investigators propose two phases of research. In phase 1 they will conduct an experiment to improve understanding of the patient roles, expectations, and perceptions that may mold interview mode differences. They will also conduct a feasibility study of a new touch screen ACASI system for use in a clinic population with low levels of education. They will also conduct a laboratory experiment to define the performance characteristics of a new DNA-based assay to detect sperm in vaginal fluid. While Phase 1 findings may be interesting in and of themselves, they have been designed to guide the implementation and evaluation of a large Phase 2 clinic-based study. In Phase 2, they will collect data and biologic specimens (vaginal swabs and urine) from 1800 STD clinic patients, half of whom will be new patients and half will be repeaters. Depending on Phase 1 findings, half of respondents will be randomly assigned to either touch screen ACASI or standard keyboard ACASI; the other half will be assigned to IAQs. The data from this study will be used to estimate interview mode effects on prevalence estimates of HIV risk-associated behaviors. They will use the new sperm assay plus urine-based assays for gonorrhea and chlamydial infection to validate measures of recent unprotected intercourse.
| Rogers, S M; Willis, G; Al-Tayyib, A et al. (2005) Audio computer assisted interviewing to measure HIV risk behaviours in a clinic population. Sex Transm Infect 81:501-7 |
| Van Lenten, B J; Wagner, A C; Nayak, D P et al. (2001) High-density lipoprotein loses its anti-inflammatory properties during acute influenza a infection. Circulation 103:2283-8 |
